Hello, Jiang. I spoke with one of the engineers at TargetScan concerning the non-conserved prediction results and he had the following to say:
"We've actually never linked the rest of our miRNA target predictions to genome coordinates because the lower-confidence predictions (any site that isn't conserved or belongs to a nonconserved miRNA family) are two orders of magnitude more numerous and the sites that our models predict are most likely to be functional (conserved sites of conserved miRNA families) get lost in the noise. Summarizing each site with just one score makes it impossible to capture all of the associated prediction metrics. Also, linking sites using UTR coordinates to genome-based coordinates is trickier than it seems like it should be, largely because of small gaps in multiz alignments that require iterative map-and-check to handle off-by-several bugs. Instead, people can download lots of information about all sites from pages like http://www.targetscan.org/cgi-bin/targetscan/data_download.cgi?db=vert_60 although it is limited in not including genome positions." Please contact us again at [email protected] if you have any further questions. --- Steve Heitner UCSC Genome Bioinformatics Group -----Original Message----- From: [email protected] [mailto:[email protected]] On Behalf Of chunjiang he Sent: Wednesday, November 30, 2011 9:48 AM To: [email protected] Subject: [Genome] a question about retrieving coordinates of targetscan target sites Dear Mr/Ms, I want to ask a question about targetscan target sites. I can get the genomic coordinates of targetscan target sites from the table track browser. But there are only about 50,000 hits and only in conserved prediction result. I see in targetscan website, there have conserved and non-conserved prediction result for more than millions of hits. The target sites are relative to 3'UTR sequence, but not correspond to genome. So, I want to ask if there is any method that I can transfer these relative positions to genomic coordinates by ucsc table browser or some others. Thanks very much, Jiang _______________________________________________ Genome maillist - [email protected] https://lists.soe.ucsc.edu/mailman/listinfo/genome _______________________________________________ Genome maillist - [email protected] https://lists.soe.ucsc.edu/mailman/listinfo/genome
